Table of contents

Some inequalities for the Cowley parameters describing atomic short-range order in substitutional alloys are derived and discussed.

An internal-friction peak of general occurrence in superconductors, centred approximately on T_c, and about 0.1-1.0K wide, independent of frequency and amplitude, is shown to arise as a result of the enhancement of the background internal friction through the association of vibrating dislocations with superconducting filaments.

Stationary regimes of the motion of phase boundaries accompanied by the heat release have been investigated. The heat produced can accelerate the phase boundary motion which in many two-phase systems is thermally activated (an example of practical importance being the grain boundary motion in the process of recrystallisation). This in turn increases the heat production rate. A system involved in this avalanche-like process can be stabilised by the heat removal via heat conductivity and heat exchange with the thermal bath establishing a stationary regime of the phase boundary motion. Both the temperature distribution around a stationarily moving phase boundary and its velocity have been analysed self-consistently. The conditions for the existence of one or two stationary regimes of the phase boundary motion have been obtained and the possibility of transitions between the two regimes accompanied by a jump-like change of the phase boundary velocity has been discussed.

The lifetimes and energies of the electronic excitations have been calculated in copper, using the standard augmented plane wave (APW) basis. The results show that the inverse lifetimes are slightly band-dependent and follow practically a parabola in the full energy range the author studied.

The bandstructure, density of states, Fermi surface and external DHVA areas are presented for ZrZn₂ and TiBe₂. Arguments are given to suggest the feasibility of DHVA experiments.

The field dependence of the susceptibility of some spin glass materials is interpreted in terms of Neel's model of superparamagnetism, leading to a universal plot for the following substances: (La,Gd)Al₂, MnO.Al₂O₃.SiO₂, Cu-Mn, Co-Ga, Eu-Sr-S. Values of the mean zero field blocking temperature and the mean anisotropy field H_K are deduced, the latter also by an independent method. The origins of H_K are discussed.

The authors have used the thermal activation model for polarisation domains in antiferromagnetic chromium to predict the ultrasonic attenuation in the magnetic field-temperature plane, explaining some experimental ultrasonic data and making predictions of new phase boundaries.

Relaxed structural models are used to present the construction and structural analysis of the amorphous metallic alloy Fe_100-xP_x. The pair distribution functions agree fairly well with experimental values. The analysis of the Voronoi polyhedra can be understood in terms of the changes in alloy composition and the tightly captured P atoms in the framework of the Fe atoms. Analysis of the geometrical configuration of near-neighbour Fe atoms around P atoms shows that the symmetry of the local atomic environment around the P atoms is very similar to that in the crystalline Fe₃P.

The authors measured at 20K the magnetoresistance curves of two high-purity polycrystalline iron samples in small longitudinal magnetic fields (H/sub ///<1000 G): (i) during a 2.5 MeV electron irradiation, (ii) during the recovery of the damage. They show that the resistivity anisotropy of their samples ( delta rho = rho /sub ///- rho _{perpendicular to} ) is closely connected to their residual resistivity rho ₀. During the irradiation (increase of rho ₀ due to the increase of scattering centres) the authors go gradually from a region of negative delta rho (the 'normal' magnetoresistance is predominant) to a region of positive delta rho in which the spontaneous anisotropy dominates. The authors can deduce from their measurements the value delta rho / rho /sub ///=0.3% of the spontaneous anisotropy relative to the diffusion by Frenkel pairs.

A theoretical study is made of the lattice stability of Ni subjected to various kinds of simple stresses. Calculations are carried out for the Ni crystal using two-body exponentially attractive and exponentially repulsive potentials, and also for two-directional stresses of equal magnitude acting simultaneously along the a₂ and a₃ directions of the crystal. The region of lattice stability in the a₂=a₃ plane of the a₁, a₂, a₃ space is also investigated using the generalised Morse functions phi _m(r)=(D/(m-1))(exp(-m alpha (r-r₀))-m exp(- alpha (r-r₀))) taking two extreme values of m, namely 1.25 and 6. Cases of (i) unidirectional stress, (ii) two-directional stresses, (iii) unidirectional deformation and (iv) application of hydrostatic pressure are shown to arise as specific cases of the generalised calculations presented here. The possibility of phase transition from FCC phase to BCC phase is discussed.

Inconsistencies between published data on breakdown orbits and those on primary orbits in metallic (white) tin are examined. A value for the area of the epsilon ₁⁵ primary orbit of 0.3035+or-0.0013 au is deduced.

In an alloy, the energies of the ionic core states in the full crystal potential cannot be determined without minimising the total energy. Thus in a priori calculations there is an indeterminacy in the ionic pseudopotentials. How this arises is demonstrated both in the conventional pseudopotential scheme and in the formalism of the screened uniform charge model. The consequences of this indeterminacy and their size are shown by calculating the resistivity of liquid lithium-magnesium alloys at various concentrations.

The resistivity of metals with non-magnetic impurities is calculated using the microscopic Fermi liquid theory based on the Keldysh temperature technique. Such low temperatures are considered that the quasiclassical description of the kinetic processes is inapplicable, i.e. T<c theta_D for the electron-phonon interaction and T<cE_F for the Coulomb electron-electron interaction (where c denotes the concentration of the defects). The temperature dependence of the resistivity in the temperature region considered arises from the electron-electron interaction interfering with the impurity scattering. Such interference results in the diffusion amplification of the two-particle interaction amplitudes. The electron-phonon interaction gives a negligibly small contribution to the resistivity when T<c theta , as compared with that of the Coulomb interaction, since the scattering by 'pure' phonons is largely compensated by that of the vibrating impurities. The Coulomb interaction between the electrons which are scattered by static impurities results in decreasing resistivity versus increasing temperature according to Delta rho (T) approximately -c^5/2(T/E_F)^1/2; for T>or approximately=c theta_D(theta_D/E_F)^1/3 this law is replaced by the dependence Delta rho (T) approximately c(T/ theta _D)², i.e. the curve Delta rho (T) possesses a minimum.

The surface magnetisation is determined for a semi-infinite simple cubic crystal described by a one-band Hubbard model. The author assumes changes in the model parameters on the first four layers of the crystal and he determines the magnetic moments for those layers by solving the eight Hartree-Fock equations numerically for the number of particles of each spin direction. The local densities of states required in the Hartree-Fock equations are determined by an integration over the Green functions for the semi-infinite crystal. Starting directly from the definition of the Green function as the resolvent of the Hamilton operator and proceeding by direct inversion of the corresponding infinite matrix, he derives an algorithm which significantly shortens the computation of the Green functions since it does not require the solution of a system of Dyson equations as in the Kalkstein-Soven method. The results of the author's numerical calculations show how the magnetic moments of the first four layers are reduced by a decrease of the Coulomb matrix elements or a downward shift of the bands and how they are enhanced by decreasing hopping matrix elements. Within the author's model calculation, he shows under what conditions dead layers could appear and how rapidly the magnetisation reaches its bulk value. The influence of hydrogen adsorption on the surface magnetisation of a model ferromagnet is briefly discussed.

A method proposed in 1971 by Meyer, Umar and Young for non-transition metals, which combined the density functional theory and the ionic pseudopotential concept was used here to study the two pure metal parameters relevant for alloy formation: the electrochemical potential and the electron density at the boundary of atomic cells. The method was improved by introducing correlation. The inhomogeneity of the density was found to influence the value of the chemical potential. A linear relation was discovered between the cell-boundary density and the dipole barrier at a metallic surface. This relation was used to set up a simple model for the variation of the work function with the temperature.

Starting from an effective Fuchs-Sondheimer conduction model, the theoretical expression for the thermoelectric power of polycrystalline metal films is derived. From the approximate expression for thick films, a procedure is proposed to determine the variation in the electronic mean free path.

The polarization analysis technique is used to obtain an unambiguous separation of the magnetic contribution to diffuse neutron scattering by MnSi from other contributions. The paramagnetic response function obtained from the results is peaked in q at around 2.5 AA^-1 and has an energy width greater than 15 THz. The absolute magnitude of the response at ambient temperature is consistent with the effective moment of 2.2 mu B per manganese atom obtained from susceptibility measurements but it decreases with temperature, although a significant amount of diffuse magnetic scattering is still in the ordered phase at 9K. These results cannot be explained quantitatively in terms of either localised or itinerant models of metallic magnetism.

The studies performed on ferrimagnetic Gd(Fe_xAl_1-x)₂ compounds, with x>or=0.4, show a small deviation of the iron magnetisation from the simple dilution behaviour. This is ascribed to the reduced exchange splitting of iron 3d states as the iron is substituted by aluminium. The reciprocal susceptibility follows a Neel-type hyperbolic law. The effective iron moments decrease as the aluminium content increases. The ferrimagnetic resonance measurements show that the g_eff values can be described by using the Wangsness' relation with g_Gd=2.00 and g_Fe=(2.08+or-0.3). The electron paramagnetic resonance studies show that the g values and the slope of the linewidth are composition and temperature dependent. The data were analysed in a model based on the molecular field approximation. Finally, the magnetic behaviour of iron in these compounds is discussed.

The bandstructure of antiferromagnetic gamma -manganese was recently calculated using the local spin functional approach and assuming, as is usual, as simple collinear spin configuration. This paper investigates the stability of this spin configuration with respect to the formation of a multiple spin density wave corresponding to a tetrahedral spin arrangement which also agrees with present experiments. The calculations which use the linear response theory within the local spin density functional approach show that the collinear spin arrangement is indeed unstable, thus favouring the multiple spin density wave as the ground-state configuration of antiferromagnetic gamma -Mn.

A theoretical model for the electromagnetic generation and detection of ultrasound in a ferromagnetic metal is described and developed. The model shows how the normal (Lorentz) processes are enhanced by a strong magnetoelastic coupling. It is used to deduce quantitative values for both the efficiency of generation and the attenuation coefficient of a transverse sound wave propagating parallel to an applied magnetic field in a single-domain sample of ferromagnetic nickel. The extension of the model to an unsaturated, multidomain sample is discussed.

Ferromagnetic resonance measurements are reported for amorphous alloys with composition (Fe_1-xNi_x)₈₃Bi₁₇, where x=0, 0.2, 0.4, 0.5 and 0.6. The measurement frequency was 9.2 GHz. The resonance linewidth narrowed from 0.035 to 0.012 T at room temperature as x varied from zero to 0.6. Hence inhomogeneity on an atomic scale is not an important source of linewidth. As the measurement temperature was reduced from 300K to 90K there was a small line broadening. Inhomogeneity in either the demagnetising field of the magnetoelastic energy may cause line broadening as a function of x and T. No significant effect upon the linewidth was detected when the specimens were annealed.

Nuclear quadrupole effects in the ⁵¹V and ⁵⁹Co NMR in sigma phase V-Co and V-Ni are studied for various alloying concentrations. Both ⁵¹V and ⁵⁹Co spectra in V-Co are used to compare the quadrupole moments of the two nuclei and to reconfirm the ratio mod Q(⁵⁹Co) mod / mod (⁵¹V) mod approximately=10. Local-environment effects are studied in quadrupole splittings of ⁵¹V resonance in both V-Co and V-Ni, and are compared with the calculated values taking into account only atoms on the coordinating polyhedron around the specific atomic site. The changes in the asymmetry parameter eta of the electric-field gradient at certain atomic sites sensitively reflect the local arrangement of atoms. The experimental values of e²qQ/h and eta for ⁵¹V are 2.6 MHz and about 0 for site II, and 2.2 MHz and about 1 for sites III and V.

Nuclear spin echo spectra are observed for three phases ( alpha _H, alpha _I, alpha _L) of CoV alloys with 11.4-21.8 at.% V at 4.2K in zero external field. The ⁵⁹Co spectra have three peaks and the hyperfine fields corresponding to the peak frequencies at 15.0 at.% V are: 136, 155, 174 kOe for the alpha _H phase; 137, 165 195 kOe for the alpha _I phase; and 141, 168, 201 kOe for the alpha _L phase. On the reasonable assumption that Co atoms with more than four vanadium atoms at their nearest-neighbour (NN) sites do not have magnetic moments, the hyperfine fields obtained in a given phase are attributed, in order from the largest value, to Co atoms with one, two and three vanadium atoms at the NN sites, respectively. These Co atoms with one to three V neighbours seem to have well defined localised moments that do not differ very much from those in Co metal.

Nuclear magnetic resonance (NMR) has been used to determine the activation energies for hydrogen diffusion in ZrH_1.75, ZrH_1.96, Zircaloy-2-H_1.90 and Zr_0.67Rh_0.33H_0.95. The proton spin-lattice relaxation time, T₁, was determined at a frequency of 60.15 MHz as a function of temperature over the range 200-900K. For ZrH_1.75 the minimum value of T₁ was found to occur at 680K whereas for the others the T₁ minima all occurred at 850K. The activation energies found were 0.55 eV for ZrH_1.75, 0.80 eV for ZrH_1.96, 0.86 eV for Zircaloy-2-H_1.90 and 0.80 eV for Zr_0.67Rh_0.33H_0.95 from the T₁ data near the T₁ minima. Using the low-temperature T₁ data, where the relaxation rate due to diffusion could be neglected, values of the electronic contribution to the spin-lattice relaxation time (T_1e) were obtained. The quantity (G_1eT)^-1/2 was found to increase with temperature, T, over the range 200-500K.

The temperature dependence of the ⁵⁷Fe Mossbauer f factor, center shift and linewidth of ⁵⁷Co:PdH_x sources with x=0 and 0.67 and of ⁵⁷Fe:PdH_x absorbers with x=0.55 and 0.71 were measured between 100 and 300K. In all hydrides an anomalous decrease of the f factor was observed between 200 and 260K. This can be attributed to the transient strains produced by the hydrogen atoms diffusing in the Pd lattice. In the samples with x=0.67 and 0.71, the temperature dependence of the centre shift deviates from the mere second-order Doppler behavior. This reveals changes in the mean population of the hydrogen sites near the Mossbauer probe. The temperature dependence of the linewidth can be understood qualitatively as a superposition of two effects, namely broadening due to the phase incoherence that arises from the fluctuating strains and motional averaging of the hyperfine interactions, which cause an inhomogeneous line broadening at low temperatures.